Invasive plant probability prediction outputs and code for paper "Modelling species distributions and environmental suitability highlights risk of plant invasions in western United States"

This data publication contains predictive maps of probability of presence of 15 invasive species across the arid and semi-arid western United States, and custom R functions used to generate them from downloaded publicly available survey data. Input data were collected 1940-2020, with 50% of the data collected since 2002. We modeled species distribution using presence/absence data from > 148,000 vegetation survey plots to predict the probability of presence of each species based on associated climate, soil, topography, and disturbance records. Methods for deriving these predictions and caveats for their use are described in McMahon et al. (2021), with summary information included in this data package. This package includes two rasters of predicted probability of occurrence per species, one representing the area within a buffer distance to known presence of the species where the model was trained and evaluated, and one covering the arid and semi-arid Environmental Protection Agency Level III ecoregions where the species was observed (lower confidence outside model training area).
The purpose of this work was to define the environmental niches and predict potential invasion risk by 15 plant species of concern in the arid and semi-arid western United States. The model predictions of invasive species presence are supplied in raster format and with associated code to facilitate their use in land management and further research, as a supplement to the published journal article.
For complete information about these data please read the associated article, "Modelling species distributions and environmental suitability highlights risk of plant invasions in western United States" (McMahon et al. 2021) before using these model outputs.

Important Note: These model outputs are intended as a resource for prioritizing monitoring and management at moderate to broad spatial scales (e.g. subwatershed) in conjunction with local knowledge and field verification. Predictions of species presence outside the model training areas represent hypotheses for where the species may occur next.

Metadata were originally published on 02/01/2021. Minor metadata updates were made on 03/15/2021.